search

harryjo97 / GDSS

Official Code Repository for the paper "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations" (ICML 2022)
https://arxiv.org/abs/2202.02514
139 stars 22 forks source link

num_workers > 0 #13

Open pavelbatyr opened 1 year ago

pavelbatyr commented 1 year ago

Hi, and thanks for the code!

I suggest using num_workers > 0 in the dataloaders. It can speed up training significantly. Using 4 workers makes training 30% faster in my case.

harryjo97 commented 1 year ago

Hi,

Thank you for the suggestion! We would add the num_workers feature to the following line. https://github.com/harryjo97/GDSS/blob/4d96334fd0d07577f9891e9d5e81dae4d64a92fd/utils/data_loader.py#L12

pavelbatyr commented 1 year ago

Thanks! There's also a dataloader for molecules in utils/data_loader_mol.py